for me 5 min chart, wait for pull back and then go in. if it does not pull back stay out, even if it goes on to rally thats ok, not part of my strategy...ignorant and points left on the table maybe, but i don't chase the market. great thread by the way!
djx I have to run as I am done trading for the day and have some errands to run and if the post the number of points I made today, ripper will got bezerk. But I a will answer all questions as soon as I can.
has anyone heard the cbot raised echange rates 20 cents rountrip per contract for the emini? has anyones broker raised there ym commissions?
Volente got me thinking when he mentioned a bias based on day of the week, so I did a data study on it. Here are the results of a study I did for the Dow. This is not based on futures YM data, but on the Dow itself, symbol $INDU. It's a simple study which attempts to correlate day of week with whether it was an "up" day or a "down" day. I used daily OHLC data going back to late 2000. For example, the number of Tuesdays was counted, along with the total number of "up" and "down" Tuesdays. These counts were tallied for each day of the week. I did the "up" and "down" two different ways. One is using the close-close relationship, the other is the open-close. Close-close is the relationship of a particular day's close with the previous trading day's close. Open-close is the relationship between a particular day's close to its open. So today is an "up" day for close-close if today's close is greater than yesterday's close. Today is an "up" day for open-close if today's close is greater than today's open. The posted results show the counts for the various combinations. I then computed something I called the "edge". The edge is first calculated by taking the number of "up" days divided by total days for that day of the week. Then I subtracted 0.5 from this number. The idea behind this is that we're looking for a real edge over coin flipping, so we want the probability to be greater or less than 50%. The more positive the edge, the higher the probability of an "up" day. The more negative the edge, the higher the probability of a "down" day. Values close to 50% are considered worthless (flip a coin). Here are the results summarized: Monday: weak sell Tuesday: moderate buy Wednesday: weak sell Thursday: strong buy Friday: moderate buy Check out the attached .xls file for details.
Fun thread this, thanks folks. Looks like volente's theory has been borne out on the above evidence, except for Friday. Interesting. Some random YM stuff from today... (twin monitor screenie, so you need to turn off automatic image resizing).
I've also done some studies on up-down patterns using 1, 2, 3, 4, and 5 day patterns, if anyone's interested. Today has a 60%+ probability of being an open-close "up" day.
steve, Great study there! I was just thinking about this myself. and thanks for sharing. What do you mean about 1-5 pattern days? do you mean 2 up days = more likely a 3rd up day? that type?
Yep, exactly. I got the ideas for the up-down pattern approach from Toby Crabel's book about Opening Range Breakout. Then I wrote a program to do it.